ABSTRACT
Over the past decades, advancements in image acquisition systems, computational resources, and algorithmic designs have enhanced the potential use of artificial intelligence in various breast cancer imaging tasks. Breast image interpretation is an important task of radiologists in both breast cancer screening and diagnostic work-up to ensure optimal patient management. In addition, breast imaging and associated analyses have the potential for use in risk assessment, prognosis, response to therapy, and risk of recurrence, as well as multi-omics cancer discovery. Clinical image acquisition systems include full-field mammography and tomosynthesis, breast MRI, and breast ultrasound, with various modifications providing specific morphological or functional characteristics of cancer. In addition, integration of information from multiple patient tests such as clinical, molecular, imaging, and genetic testing is expected to yield improved cancer diagnosis and treatment management. Thus, with regards to breast imaging, there are multiple efforts to convert image data to quantitative data through radiomics and radiogenomics. This chapter touches on the various areas of rapid development in breast image analysis due to the continuous rise in machine learning, including deep learning.
